I


THESIS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY

Nanostructures of Graphite and Amorphous Carbon - Fabrication and Properties

Hans Fredriksson

Department of Applied Physics CHALMERS UNIVERSITY OF TECHNOLOGY

Göteborg, Sweden 2009


 II


Nanostructures
of
Graphite
and
Amorphous
Carbon
‐
Fabrication
and
Properties


©
HANS
FREDRIKSSON,
2009

ISBN
978‐91‐7385‐307‐1


Doktorsavhandlingar
vid
Chalmers
tekniska
högskola
Ny
serie
nr
2988
ISSN
0346‐718X


Department
of
Applied
Physics
Chalmers
University
of
Technology
SE‐412
96
Göteborg
Sweden
Telephone
+
46
(0)31‐772
1000


Cover:
Optical
extinction
spectra
from
amorphous
carbon
nanostructures
with
decreasing
sizes
and
SEM
images
of
nanostructured
graphite
samples
with
three
different
feature
sizes.














Printed
at
Chalmers
Reproservice

Göteborg,
Sweden
2009



 III


ABSTRACT
Nanoscience is a well-established research area, which concerns properties and fabrication of objects with typical dimensions on the 1-100 nanometer length scale. A central issue has been the development of techniques for fabrication and characterization of nanometer sized objects, which have contributed considerably to progress in both practical applications and fundamental research. Still, a standing challenge in nanofabrication is to further decrease the size limit and increase the precision in structure fabrication, with a simultaneous increase in reliability and cost-efficiency. Other goals are to facilitate fabrication of nanostructures in a variety of materials, with different geometries and spatial distributions. Examples of practical applications of nanofabrication are, electronic devices, nanoparticle reinforced composite materials, materials for extraction and storage of energy, sensors and biomedical applications. In this thesis, the development and application of a nanofabrication technique termed hole-mask colloidal lithography (HCL) is described. The technique is based on self-assembly of nanospheres in combination with spin coating and thin film evaporation to produce supported nanostructured masks for etch and/or deposition processes. HCL relies on a parallel process and uses relatively simple laboratory equipment. Therefore it is fast and cost-effective and can be used to structure large surface areas in a reasonable time. Furthermore, HCL is suited for fabrication of nanostructures with a variety of different shapes, with well-defined sizes and in a large variety of different materials. Demonstrated examples include discs, ellipses, bi-metallic particle pairs, cones and inverted ring structures in Au, Ag, Cr. Specifically, the use of HCL to fabricate nanostructures in three different carbon materials, highly oriented pyrolytic graphite (HOPG), glassy carbon (GC) and amorphous carbon, is described. Such nanostructured materials are relevant both in technical applications and in model studies of e.g. soot particles. The manufactured nanostrucutres have been characterized with respect to their geometrical, mechanical, and optical properties, using microscopy and spectroscopy techniques, and their reactivity towards oxidation has been explored. From studies of such samples, it is concluded that the etch rate in oxygen plasma is different for HOPG and GC, which influences the resulting size and shape of the nanostructures after the applied oxidation treatment. It is also shown that the atomic arrangement of the HOPG nanostructures is similar to that of the bulk material. Investigations of the optical properties reveal resonant absorption and scattering of light for nanostructures in all three materials, i.e. peak position, amplitude and width of the measured optical spectra are shown to correlate with the nanostructure sizes. This correlation is used to optically monitor oxidation, and the resulting decrease in volume, of carbon nanostructures under high temperature oxidation conditions and is proposed as a general sensing method to study oxidation/combustion of soot and other carbon nanostructures. Keywords: Nanofabrication, carbon, HOPG, GC amorphous carbon, oxidation, optical resonance, optical spectrum, Raman spectrum.


 IV



 V



 VI


LIST OF PUBLICATIONS This thesis is based on the work contained in the following papers: Paper I. Hole-mask colloidal lithography

H. Fredriksson, Y. Alaverdyan, A. Dmitriev, C. Langhammer, D. S. Sutherland, M. Zäch and B Kasemo Advanced Materials, 2007. 19 (23): p. 4297

Paper II. Patterning of highly oriented pyrolytic graphite and glassy

carbon surfaces by nanolithography and oxygen plasma etching H. Fredriksson, D, Chakarov and B. Kasemo Carbon, 2009. 47: p. 1335

Paper III. Resonant optical absorption in graphite nanostructures

H. Fredriksson, T. Pakizeh, M. Käll, B. Kasemo and D. Chakarov Journal of Optics A; Pure and applied Optics, 2009. 11, in press

Paper IV. Raman spectroscopy of nanostructured graphite

H. Fredriksson, J. Cardenas, B. Kasemo and D. Chakarov Submitted to Nanotechnology

Paper V. Oxidation of lithographically prepared amorphous carbon

soot models monitored by optical spectroscopy H. Fredriksson, T. Pakizeh, D. Chakarov and B. Kasemo Manuscript in preparation

Peer reviewed paper, produced during the thesis period, that is not included in this thesis: Enhanced nanoplasmonic optical sensors with reduced substrate effect

A. Dmitriev, C. Hägglund, S. Chen, H. Fredriksson, T. Pakizeh, M. Käll and D. S. Sutherland Nano Letters, 2008. 8 (11): p. 3893


 VII


My contribution to the papers included in the thesis: Paper I. Contributed
with
the
idea
and
developed
the
fabrication


technique
after
discussions
with
Bengt
Kasemo.
Performed
most
of
the
experimental
fabrication
work.
Principal
author
of
the
paper.



Paper II. Contributed
with
all
the
experimental
work
and
data
analysis.
Principal
author
of
the
paper.



Paper III. Contributed
with
all
the
experimental
work
and
the


major
part
of
the
data
analysis.
Principal
author
of
the
paper.



Paper IV. Contributed
with
all
fabrication
work,
part
of
the
optical


measurements
and
the
major
part
of
the
data
analysis.
Principal
author
of
the
paper.



Paper V. Contributed
with
all
the
experimental
work
and
the
major


part
of
the
data
analysis.
Principal
author
of
the
paper.




 VIII





Table
of
Contents
1
 Introduction .................................................................................................................1
1.1
 Nanoscience....................................................................................................................... 1
1.2
 Applied
nanoscience ...................................................................................................... 2
1.3
 Carbon
nanostructures.................................................................................................. 4
1.4
 Scope
and
motivation
of
the
thesis:........................................................................... 5
1.4.1
 Development
of
a
nanofabrication
technique ............................................................. 5
1.4.2
 Fabrication
of
carbon
nanostructures ............................................................................ 6
1.4.3
 Investigation
of
carbon
nanostructure
properties .................................................... 6


2
 Nanofabrication ..........................................................................................................9
2.1
 Pattern
writing
techniques .......................................................................................... 9
2.2
 Pattern
replicating
techniques .................................................................................10
2.3
 Selfassembly
techniques ...........................................................................................10
2.4
 Challenges
and
limiting
factors ................................................................................11


3
 Holemask
colloidal
lithography
(HCL)............................................................ 13
3.1
 Colloidal
lithography....................................................................................................13
3.2
 Sparse
colloidal
lithography......................................................................................13
3.3
 Holemask
colloidal
lithography..............................................................................15
3.3.1
 Fabrication
of
a
supported,
patterned
mask..............................................................15
3.3.2
 Transfer
of
the
mask
pattern
through
reactive
ion
etching ................................17
3.3.3
 Transfer
of
the
mask
pattern
through
deposition ...................................................20
3.3.4
 Limitations
and
extensions................................................................................................23


3.4
 Applications
of
HCL
to
fabricate
carbon
nanostructures.................................24
3.4.1
 Carbon
nanostructures
from
bulk
materials..............................................................25
3.4.2
 Evaporated
carbon
nanostructures ...............................................................................26


3.5
 Other
applications
of
HCL...........................................................................................27
4
 Carbon ......................................................................................................................... 29
4.1
 Atomistic
origin
of
different
phases........................................................................29
4.2
 Crystalline
graphite ......................................................................................................31
4.2.1
 Crystal
structure.....................................................................................................................31
4.2.2
 Band
structure.........................................................................................................................33


4.3
 Graphitic
carbon
materials ........................................................................................34
4.3.1
 Crystalline
graphite,
synthetic
and
natural ................................................................35
4.3.2
 Glassy
carbon...........................................................................................................................35
4.3.3
 Graphitic
amorphous
carbon ............................................................................................36


4.4
 Optical
properties
of
crystalline
graphite.............................................................36
4.4.1
 Parallel
to
the
ab‐plane........................................................................................................36
4.4.2
 Perpendicular
to
the
ab‐plane..........................................................................................38


4.5
 Optical
properties
of
glassy
carbon ........................................................................39
4.6
 Optical
properties
of
amorphous
carbon..............................................................39
4.7
 Raman
spectrum
of
crystalline
graphite ...............................................................39
4.8
 Raman
spectrum
of
glassy
carbon...........................................................................44
4.9
 Raman
spectrum
of
amorphous
carbon ................................................................44
4.10
 Oxidation
of
crystalline
graphite...........................................................................45
4.10.1
 Experimental
observations .............................................................................................46
4.10.2
 Theoretical
considerations .............................................................................................46


4.11
 Oxidation
of
glassy
carbon.......................................................................................48



 IX


4.12
 Oxidation
of
amorphous
carbon ............................................................................48
5
 Carbon
Nanostructures ......................................................................................... 51
5.1
 Different
types
of
carbon
nanostructures.............................................................51
5.2
 Optical
properties
of
nanostructures .....................................................................53
5.3
 Optical
properties
of
carbon
nanostructures ......................................................54
5.4
 Raman
scattering
by
nanostructures .....................................................................56
5.5
 Raman
scattering
by
carbon
nanostructures ......................................................58
5.6
 Oxidation
of
carbon
nanostructures.......................................................................59
5.6.1
 Experimental
observations................................................................................................59
5.6.2
 Theoretical
considerations ................................................................................................60


6
 Experimental............................................................................................................. 63
6.1
 Oxidation
systems .........................................................................................................63
6.1.1
 Reactive
Ion
Etching,
RF
glow
discharges ...................................................................63
6.1.2
 Gas‐flow
reactor .....................................................................................................................64


6.2
 Optical
characterization ..............................................................................................64
6.2.1
 On‐line
optical
extinction
spectroscopy.......................................................................64
6.2.2
 Off‐line
optical
extinction
spectroscopy ......................................................................66
6.2.3
 Raman
spectroscopy.............................................................................................................66


7
 Results
and
discussion........................................................................................... 69
7.1
 Summary
of
papers.......................................................................................................70
7.1.1
 Paper
I.
Hole‐mask
colloidal
lithography ....................................................................70
7.1.2
 Paper
II.
Patterning
of
highly
oriented
pyrolytic
graphite
and
glassy


 carbon
surfaces
by
nanolithography
and
oxygen
plasma
etching....................72
7.1.3
 Paper
III.
Resonant
optical
absorption
in
graphite
nanostructures ................73
7.1.4
 Paper
IV.
Raman
spectroscopy
of
nanostructured
graphite ...............................74
7.1.5
 Paper
V.
Oxidation
of
lithographically
prepared
amorphous
carbon


 soot‐models,
monitored
by
optical
spectroscopy....................................................75


7.2
 Discussion ........................................................................................................................76
7.3
 Outlook .............................................................................................................................77


8
 Acknowledgements................................................................................................. 79


9
 References.................................................................................................................. 81




 X







 1





1 Introduction
1.1 Nanoscience

Since
 the
 dawn
 of
 time,
 by
 curiosity
 as
well
 as
 for
 practical
 reasons,
man
 has
been
driven
to
explore
and
to
look
closer
at
nature.
Especially
the
extremes
have
always
 fascinated;
 the
 extremely
 large,
 distant,
 strong
 or
 small.
 An
 impressive
amount
 of
 creativity
 and
 thought
 have
 been
 invested
 into
 developing
instruments
 and
 aids
 to
 facilitate
 investigations
 beyond
 the
 capabilities
 of
 our
senses.


For
the
study
of
objects
that
are
too
small
to
apprehend,
so
small
that
the
human
eye
 cannot
 resolve
 their
 position
 and
 shape,
 a
 multitude
 of
 tools
 have
 been
constructed.
From
simple
magnifying
glasses,
via
optical
microscopes,
to
today’s
electron‐
 and
 scanning
 probe
 microscopes,
 with
 which
 single
 atoms
 can
 be
observed.
The
access
to
such
characterization
tools
has
opened
up
for
a
research
field
dedicated
to
study
nature
on
an
extremely
small
scale.


The
study
of
objects
with
sizes
conveniently
expressed
in
nanometers
(10‐9
m
or
equivalently
 one
 millionth
 of
 a
 millimeter)
 is
 termed
 nanoscience.
 This
 size
regime
 concerns
objects
with
 sizes
 from
1
 ‐
 1000
nm
 (often
 the
upper
 limit
 is
given
 as
 100
 nm).
 Objects
 with
 these
 characteristic
 sizes
 have
 always
 been
present
 in
 our
 environment.
 For
 example,
 biological
 nanoscale
 objects
 such
 as
viruses
and
proteins
as
well
as
inorganic
aerosols
have
sizes
in
this
regime
and
already
in
the
mid
19th
century
colloidal
solutions
containing
nanospheres
were
prepared
intentionally
for
scientific
purposes
[1].

For
 esthetical
 and
 practical
 reasons
 as
well
 as
 for
 amusement,
miniaturization
has
also
been
a
long‐standing
fascination
for
mankind.
From
a
practical
point
of
view
 it
 is
 obvious
 that
 scaling
 down
 size
 of
 useful
 gadgets
 can
 be
 very
convenient.
 An
 example
 from
 the
 past
 is
 manufacturing
 of
 portable
 wrist‐
 or
pocket‐
 watches,
 which
 are
 clearly
 more
 convenient
 to
 bring
 along
 than
 a
standard
 wall
 clock.
 A
 later
 and
 even
 more
 elucidating
 example
 is
 the
miniaturization
of
components
used
in
computers,
which
in
the
beginning
were
large
enough
to
fill
entire
rooms
(fig.1.1).




 2




Figure
1.1
One
of
 the
 first
 computers,
 ENIAC,
weighing
30
 tons
 and
occupying
a
15x9
m
large
room,
was
built
in
the
1940s,
before
nanofabrication
techniques
were
available.


Investigations
 of
 nanosized
 objects
 have
 been
 going
 on
 in
 parallel
 with
 the
efforts
 to
 design
 and
manipulate
matter
 on
 a
 similar
 scale.
 For
 the
 purpose
 of
fabrication,
techniques
developed
to
study
nanoscale
objects
have
been
adjusted
to
also
facilitate
manufacturing.
With
both
characterization
and
fabrication
tools
capable
 of
 handling
 the
 nanometer
 size
 regime,
 it
 is
 now
 possible
 to
systematically
manufacture
devices
with
desired
nanoscale
dimensions,
examine
the
result
of
the
fabrication
process
and
to
carefully
characterize
the
nanodevice
properties.
In
this
manner,
the
influence
of
nanostructure
size
on
the
mechanical,
electrical,
 optical,
 magnetic
 and
 chemical
 properties
 (to
 name
 a
 few)
 for
 a
multitude
of
materials
have
been
investigated[2,
3].
Such
investigations
provide
valuable
 information
 about
 fundamental
 physics
 and
 contribute
 to
 the
 general
understanding
 of
 nature,
 as
well
 as
 laying
 the
 foundation
 for
 a
 variety
 of
 new
technological
products.


1.2 Applied
nanoscience

Nanoscience
 and
 fabrication
 is
 applied
 in
 a
 variety
 of
 contexts.
 Careful
investigations
 of
 functionality
 of
 catalysts,
 which
 are
 of
 enormous
 practical
importance,
 can
 be
 undertaken
 with
 a
 combination
 of
 nanofabrication
 and
characterization[4]
techniques.
Also
in
the
challenge
of
meeting
future
demands
on
sustainable
energy,
nanoscience
is
predicted
to
play
an
important
role,
e.g.
in
the
design
of
efficient
and
affordable
solar
cells
and
in
photochemical
devices
for
splitting
water
to
create
hydrogen
(fig.1.2)[5].
Another
research
field
of
current
interest,
 where
 nanoscience
 is
 of
 crucial
 importance,
 is
 climate
 modeling
 and
monitoring.
 The
 international
 panel
 on
 climate
 change
 (IPCC)
 has
 identified
aerosols
(naturally
occurring,
airborne
micro
and
nanostructures)
as
important
contributors
to
the
thermal
balance
of
earth’s
atmosphere[6].





 3


Figure
 1.2
 The
 number
 of
 publications
 using
 nanoscience
 in
 three
 different
 research
fields,
taken
from
[5].

Not
 only
 have
 nanostructures
 been
 fabricated
 and
 studied
 for
 the
 purpose
 of
investigating
 their
 physical
 properties,
 but
 fabrication
 and
 investigation
 of
nanosized
 objects
 are
 also
 of
 great
 interest
 in
 commercial
 applications.
One
 of
the
most
 highly
 developed
 commercial
 uses
 of
 nanotechnology
 is
 found
 in
 the
electronics
 industry.
 Fabrication
 of
 both
 semiconductor
 components
 and
magnetic
 storage
 devices
 relies
 on
 ‐
 and
 helps
 developing
 ‐
 nanofabrication
techniques.
 Therefore,
 computers
 and
 mobile
 electronics
 are
 examples
 of
applications
 where
 nanofabrication
 is
 the
 key
 requirement
 for
 the
 enormous
development
 that
 has
 occurred
 since
 the
 end
 of
 the
 20th
 century.
 Other
commercial
 fields
 where
 nanotechnology
 is
 applied
 are
 medicine,
 composite
material
engineering,
security
and
crime
investigations
(fig.1.3)
[7‐11].




Figure
 1.3
 Thumbs
 up
 for
 nanoscience.
 A
 fingerprint
 detected
 using
 antibody
functionalized
nanoparticles,
taken
from
[10].



 4





1.3 Carbon
nanostructures
It
is
not
only
the
size,
but
also
the
material
in
a
nanostructure,
that
determines
its
physical
 properties.
 Therefore,
 different
materials
 are
more
or
 less
well
 suited
for
 specific
 applications.
 In
 this
 thesis,
 the
material
 of
 choice
 has
 been
 carbon,
which
 is
well
motivated
 for
 a
 number
of
 reasons.
The
 importance
of
 carbon
 is
indisputable,
not
only
due
to
its
abundance
(by
number
density
the
fourth
most
abundant
element
 in
 the
universe[12]),
but
also
 for
 its
appearance
 in
 technical
applications
 as
well
 as
 for
 forming
 the
 backbone
 of
 organic
matter.
 In
 nature,
carbon
appears
in
pure
forms
as
graphite,
diamond
and
in
the
amorphous
phase,
all
 of
 which
 are
 important
 in
 technical
 applications
 such
 as
 metal
 alloys
 and
diamond
 for
 tools,
 composite
materials,
plastics,
 lubricants,
protective
 coatings
and
 high
 temperature
 applications.
 It
 also
 appears
 in
 atmospheric
 gases
 and
pollutants
 (e.g.
 CO2
 and
 soot)
 in
 fuels
 (hydrocarbons
 and
 coal)
 and
 in
 space
(interstellar
 dust).
 Furthermore,
 it
 is
 suggested
 that
 carbon
 will
 play
 an
important
 role
 in
 future
 electronics
 applications
 and
 synthetic
materials[3,
 9].
Therefore,
 the
vast
 research
efforts
 that
have
been
and
still
 are
undertaken,
 in
order
to
characterize
the
properties
of
carbon‐based
matter,
are
well
motivated.


In
addition
to
the
naturally
occurring
carbon
materials
(graphite
and
diamond)
discussed
 above,
 carbon
 also
 forms
 two
 nanostructure
 allotropes.
 One
 is
 the
Buckminster
fullerene[13]
and
the
other
is
carbon
nanotubes[14].
Since
the
early
1990s
the
single
most
studied
nanostructure
is
the
carbon
nanotube
(CNT)
and
much
 effort
 has
 been
 invested
 into
 characterizing
 this
 fascinating
 material
 as
well
 as
 into
 developing
 applications.
 Because
 of
 their
 exceptional
 mechanical
strength
 in
 combination
 with
 semi‐conducting
 or
 metallic
 as
 well
 as
 optical
properties,
CNT’s
have
been
used
 to
 fabricate
nanoelectronic
 components
 (like
transistors)[3,
 15],
 composite
 materials
 with
 high
 strength
 and
 electrical
conductivity[9]
and
efficient
light
absorbers[16].


Multitudes
 of
 other
 carbon
 nanostructures
 prevail
 and
 are
 being
 investigated
with
 similar
 and
 other
 applications
 as
 for
 the
 CNT’s
 in
 mind.
 However,
 in
addition
 to
 the
 usefulness
 of
 carbon
 nanostructures
 in
 different
 applications,
their
unintended
presence
can
cause
big
problems.
In
many
processes,
especially
combustion
of
carbon‐based
fuels
in
power
plants,
diesel
engines
and
open
fires,
soot
is
generated.
Soot
consists
predominantly
of
carbon,
forming
particles
with
sizes
 in
 the
 nanometer
 regime.
 Because
 of
 their
 abundance
 and
 impact
 on
 the
environment
and
health
 it
 is
 important
 to
know
and
understand
 their
physical
and
 optical
 properties.
 Smog
 pollutes
 and
 decreases
 visibility
 in
 our
 everyday
environment.
Soot
in
the
atmosphere
can
also
contribute
to
global
warming
via
the
 strong
 absorption
 of
 light
 (which
 eventually
 is
 turned
 into
 heat)[17‐21].
However,
 the
 contribution
 from
 carbon‐based
 aerosols
 to
 the
 atmospheres
energy
balance
is
still
under
debate
and
there
is
currently
no
consensus
on
their
net
 effect
 on
 global
 warming[6].
 Another
 aspect
 on
 airborne
 carbon
nanostructures
 is
 their
 toxicological
properties.
 Investigations
of
 these
(as
well
as
 for
 other
 nanpoparticles)
 have
 shown
 that
 negative
 health
 effects
 can
 be
expected[7,
8].
It
is
also
believed
that
carbon
nanostructures
are
responsible
for
the
optical
absorption
at
~220
nm
observed
in
astrophysical
studies[22,
23].



 5





1.4 Scope
and
motivation
of
the
thesis:
The
work
presented
in
this
thesis
is
divided
into
two
major
parts;
development
of
a
nanofabrication
technique
and
application
of
this
technique,
in
particular
for
fabrication
and
investigations
of
carbon
nanostructures.



1.4.1 Development
of
a
nanofabrication
technique
The
first
part
treats
nanofabrication
in
general
and
the
further
development
of
a
very
useful
self‐assembly
based
technique,
colloidal
lithography
(CL),
into
a
new
and
more
versatile
version
named
hole‐mask
colloidal
lithography
(HCL).
CL
is
a
relatively
 simple
and
 flexible
nanofabrication
 technique,
with
modest
demands
for
advanced
laboratory
equipment,
but
with
excellent
control
over
fundamental
nanostructure
 properties.
 Since
 it
 uses
 self‐assembling
 colloidal
 spheres
 to
define
nanostructure
size,
shape
and
spatial
distribution
it
is
also
well
suited
for
upscaling
to
patterning
of
surface
areas
of
several
tens
of
cm2.
It
is
therefore
well
suited
 both
 for
 fabrication
 of
 samples
 for
 various
 applied
 and
 fundamental
research
 projects
 and
 for
 commercially
 useful
 nanostructures.
 All
 these
advantages
 are
 shared
between
 the
 conventional
CL
 technique
 and
 the
 further
developed
version
HCL.


The
 objective
 of
 developing
 the
 HCL
 technique
 was
 to
 further
 expand
 the
applicability
 and
 versatility
 of
 CL.
 One
 of
 the
 numerous
 benefits
with
 the
HCL
technique
 is
 the
 plethora
 of
 nanostructure
 geometries
 that
 can
 be
 achieved,
using
the
same
basic
fabrication
approach.
Furthermore,
it
should
be
noted
that
the
 fabrication
 technique
 is
 essentially
 independent
of
 the
used
 substrate‐
 and
nanostructure‐
materials,
which
makes
 it
 robust
 and
easy
 to
adapt
 to
different
requirements.

Another
advantage,
which
was
also
a
direct
goal
with
developing
the
HCL
 technique,
 is
 that
no
 reactive
oxygen
 treatment
of
 the
nanostructured
samples
is
needed,
which
is
often
the
case
for
the
previous
colloidal
lithography
approaches.
This
 is
 very
useful
 for
 the
 fabrication
of
 nanostructures
of
 oxygen
sensitive
materials
such
as
carbon
and
ruthenium.




 6





1.4.2 Fabrication
of
carbon
nanostructures
The
 second
 part
 of
 the
 thesis
 concerns
 applications
 of
 the
 developed
nanofabrication
 technique
 for
 fabrication
 and
 investigation
 of
 various
nanostructures
and
their
fundamental
properties.
The
first
objective
here
was
to
demonstrate
the
versatility
of
the
HCL‐technique
by
fabrication
of
gold
and
silver
nanostructures
 with
 different
 geometries
 and
 their
 applications
 for
investigations
 of
 optical
 properties.
 Both
 gold
 and
 silver
 nanostructures
 are
known
to
exhibit
extraordinary
properties
both
for
absorption
and
scattering
of
visible
 light
and
as
substrate
in
investigations
of
adsorbed
molecules
in
surface
enhanced
Raman
spectroscopy
(SERS).
The
optical
properties
of
nanostructures
are
very
sensitive
to
their
shape
and
size.
Therefore,
a
set
goal
that
has
also
been
achieved
 was
 to
 demonstrate
 that
 HCL
 facilitates
 control
 over
 these
 relevant
properties.
A
number
of
 following
publications
also
prove
the
usefulness
of
 the
developed
 nanofabrication
 scheme
 for
 fabrication
 of
 nanostructures
 in
 other
materials
 and
 for
 other
 applications[24‐30].
 However,
 the
 bulk
 part
 of
 the
present
thesis
concerns
fabrication
and
investigation
of
carbon
nanostructures.



1.4.3 Investigation
of
carbon
nanostructure
properties
In
 this
 thesis,
 applications
of
HCL
 for
 fabrication
of
nanostructures
 in
different
types
of
carbon
materials
are
demonstrated.
Only
carbon
of
graphitic
character
was
 used,
 including
 amorphous
 graphitic
 carbon,
 glassy
 carbon
 and
 synthetic
graphite.
 The
 other
 three
 carbon
 allotropes,
 diamond,
 fullerenes
 and
 carbon
nanotubes
 were
 not
 considered.
 It
 is
 demonstrated
 that
 HCL
 can
 be
 used
 to
fabricate
 anything
 from
 crystalline
 graphite
 to
 amorphous
 carbon
nanostructures
 with
 good
 control
 over
 distribution
 and
 size.
 Samples
 with
carbon
nanostructures
covering
several
cm2
and
with
sizes
ranging
 from
a
 few
tens
up
to
several
hundreds
of
nanometers
have
been
fabricated
for
applications
in
 different
 experiments.
 These
 experiments
 investigate
 and
 compare
 the
physical
properties
of
nanostructures
in
the
different
types
of
carbon
materials,
and
correlate
these
with
structure
size
and
shape.


Two
 main
 properties
 of
 the
 fabricated
 carbon
 nanostructures
 have
 been
investigated;
their
optical
properties
and
their
reaction
with
oxygen.
The
optical
properties
were
studied
in
detail,
to
achieve
fundamental
knowledge
and
for
the
use
 in
 specific
 applications.
 Especially
 the
dependence
of
 optical
 properties
 on
nanostructure
 size,
 shape
 and
 carbon
 quality
 has
 been
 investigated.
 Optical
techniques
were
also
used
to
investigate
the
mechanical
and
thermal
properties
of
the
fabricated
nanostructures.





 7


Oxidation
 of
 carbon
 nanostructures
 is
 the
 other
 central
 property
 that
 was
thoroughly
studied.
The
response
of
different
nanostructured
carbon
materials,
both
 to
 molecular
 oxygen
 and
 to
 oxygen
 plasma,
 was
 studied.
 These
 studies
provide
information
on
the
tendency
of
oxidation
for
different
carbon
materials
under
 different
 conditions.
 Conclusions
 from
 these
 experiments
 provide
information
 that
 is
useful
 in
many
practical
 situations,
 e.g.
durability
of
 carbon
containing
materials,
combustion
of
coal,
exhaust
cleaning
and
climate
modeling.
As
a
specific
example,
combinations
of
the
results
from
the
different
studies
have
been
used
 to
 study
 combustion
of
 lithographically
 prepared,
 nanosized
 carbon
structures,
by
means
of
their
optical
properties.



 8







 9


2 Nanofabrication
A
 multitude
 of
 different
 nanofabrication
 approaches
 have
 been
 developed
throughout
 the
 years
 based
 on
 different
 fundamental
 ideas.
 This
 chapter
 will
only
provide
a
brief
discussion
of
the
different
aspects
of
the
major
techniques,
but
 several
 excellent
 reviews
 and
 books
 are
 available
 for
 the
 interested
reader[31‐34].
The
nanofabrication
methods
presented
here
can
be
divided
into
three
 major
 classes;
 pattern
 writing,
 pattern
 replicating
 and
 self‐assembly
techniques.
Each
of
these
has
individual
advantages
and
limitations.


2.1 Pattern
writing
techniques
• Electron
beam
lithography
(EBL)
• Focused
ion
beam
patterning
(FIB)
• Scanning
probe
microscopy
based
lithographies
(SPML)



These
 techniques
 are
 characterized
 by
 flexibility
 of
 the
 structure
 shapes
 and
patterns
 that
 can
 be
 produced[33‐36].
 Virtually
 any
 two‐dimensional
 pattern
within
the
resolution
 limits
can
be
“written”
onto
the
surface.
Resolution
 limits
for
 EBL
 and
 FIB
 are
 on
 the
 order
 of
 ~10
 nm
 whereas
 SPML
 techniques
 can
position
individual
atoms
and
thus
can
be
said
to
have
a
resolution
below
1
nm.
The
processes
are
serial,
i.e.
one
feature
is
written
at
a
time,
and
the
techniques
are
thus
relatively
slow
and
therefore
not
compatible
with
large
area
patterning,
although
 various
 schemes
 for
 parallel
writing
 are
 being
 developed
 to
 increase
the
 speed
 of
 these
 techniques.
 However,
 for
 research
 applications
 that
 do
 not
require
 samples
 with
 large
 surface
 areas
 and
 where
 the
 necessary
 (but
 often
expensive)
machinery
is
available,
these
techniques
are
well
suited.




 10






2.2 Pattern
replicating
techniques
• Photo
lithography
(PL)
• Nano
imprint
lithography
(NIL)
• Ion
projection
lithography
(IPL)



These
techniques
are
used
to
reproduce
predefined
patterns
from
templates
or
masks.
The
methods
are
characterized
by
high
throughput
of
structured
surface
area[34,
 35,
 37].
 Therefore,
 these
 techniques
 are
 ideal
 for
 large
 scale,
 serial
production.
 Limitations
 on
 resolution
 and
 throughput
 are
 steadily
 pushed
forward
 (resolution
 is
 currently
 well
 below
 100
 nm),
 especially
 by
 the
 semi‐conductor
 industry.
 On
 the
 other
 hand,
 the
 flexibility
 of
 these
 methods
 is
relatively
low
and
new
masks
and
templates
need
to
be
fabricated
for
every
new
pattern
 configuration.
 For
 research
 applications
 where
 nanostructure
 size,
separation
and
 shape
are
 important
variables
 these
methods
are
not
very
well
suited.
In
addition,
it
is
required
that
researchers
have
access
to
pattern
defining
equipments,
such
as
EBL.


2.3 Self‐assembly
techniques
• Polymer
self
assembly
• Colloidal
Lithography
(CL)




Patterns
 and
 structures
 are
 here
 determined
 by
 inherent
 properties
 of
 the
lithography
mask
constituents.
By
altering
the
process
parameters,
some
control
over
 the
 produced
 patterns
 can
 be
 attained.
 Therefore
 these
 methods
 are
normally
more
 flexible
 than
 the
pattern
 replicating
 techniques
but
 less
 flexible
than
 the
pattern
writing
methods[38‐40].
Simplicity
of
 the
required
equipment
and
 large
 area
 compatibility
 are
 also
 trade‐marks
 of
 the
 self‐assembly
 based
methods.
 Therefore,
 these
 methods
 are
 very
 useful
 for
 studies
 where
 sample
surface
areas
from
~1
cm2
are
required,
and
where
structural
parameters
of
the
nanostructures
are
being
investigated.
These
techniques
are
also
useful
for
large‐scale
 fabrication
 processes,
 where
 exact
 distribution
 and/or
 shape
 of
 the
nanostructures
are
not
of
central
importance.





CL
 is
 the
 category
 to
 which
 the
 fabrication
 method
 presented
 in
 this
 thesis
belongs
 and
 the
 subject
 will
 be
 treated
 in
 the
 following
 chapter[38,
 40,
 41].
Lithography
 using
 nanosize
 spheres
 to
 pattern
 surfaces
 can
 be
 seen
 as
 an
intermediate
between
the
pattern
writing
nanofabrication
methods
such
as
EBL,
FIB
 and
SPML
and
 the
pattern
 replicating
methods
PL
 and
NIL.
The
method
 is
based
on
self‐assembly
of
nanospheres
on
surfaces
and
either
the
gaps
between
close
 packed
 spheres
 or
 the
 spheres
 themselves
 are
 then
 used
 as
 an
 etch‐
 or
evaporation
 mask.
 By
 choosing
 the
 colloidal
 particle
 size,
 separation
 and
processing
conditions,
 it
 is
possible
 to
 control
 the
 size,
 separation,
distribution
and
even
shape
of
the
resulting
structures
with
an
astonishing
flexibility.





 11


Colloidal
 lithography
 can
 routinely
 produce
 structures
 with
 sizes
 down
 to
 a
couple
 of
 tens
 of
 nanometers
 and
 has
 proven
 able
 to
 fabricate
 structures
 not
easily
 achievable
 with
 EBL
 like
 sharp
 edge
 particles
 and
 hollow
 cylinders[38,
42].
 Furthermore,
 it
 is
 a
 relatively
 simple
 method
 with
 little
 demands
 for
advanced
machinery
and
it
is
a
parallel
method,
well
suited
for
patterning
large
surface
areas,
which
distinguishes
it
from
the
previously
described
lithographic
methods
(EBL,
FIB,
and
STML).
Contrary
to
the
pattern
replicating
methods
(PL,
NIL),
 CL
 requires
 no
 other
 lithographic
 technique
 to
 pre‐define
 structure
characteristics
but
possesses
intrinsic
control
over
the
pattern
parameters.


2.4 Challenges
and
limiting
factors
There
 are
 several
 challenges
 to
 be
 met
 in
 the
 further
 development
 of
nanofabrication
 methods.
 From
 a
 commercial
 point
 of
 view
 increasing
fabrication
 speed
 and
 decreasing
 structure
 size
 and
 number
 density
 at
competitive
costs,
are
the
most
central
ones.


Since
methods
 like
 STML
 can
 already
 reach
 the
 ultimate
 size
 limit
 for
 “atomic
materials”,
 the
challenge
there
is
rather
to
simplify
the
techniques
and
to
make
them
routinely
available
to
a
broader
group
of
users.
If
commercial
applications
shall
be
reached,
they
also
need
to
be
made
both
faster
and
cheaper.

FIB
and
EBL
are
already
highly
mature
techniques
capable
of
defining
structures
with
 sizes
 down
 to
 5
 nm.
 However,
 the
 serial
 nature
 of
 these
 methods
 will
probably
 restrict
 their
 broader
 applications,
 for
 quite
 some
 time,
 to
 the
fabrication
of
masks
to
be
used
in
pattern
replicating
techniques.


Developments
 in
 PL
 with
 the
 present
 techniques
 used
 commercially
 are
estimated
 to
 be
 able
 to
 reach
 down
 to
 ~45
 nm
 in
 inter‐particle
 distances.
Furthermore,
 other
 versions
 of
 photolithography
 utilizing
 extremely
 short
wavelength
 synchrotron
 radiation
 and
 interference
 gratings
 instead
 of
conventional
photolithography
masks
have
already
been
demonstrated
and
are
predicted
to
reach
even
better
resolution[37].


Imprinting
techniques
are
ultimately
limited
by
the
graininess
of
matter
but
can
theoretically
 fabricate
 structures
 down
 to
 the
 single
 nm
 regime.
 In
 practice
however
differences
 in
 thermal
expansion
properties
of
 the
 involved
materials,
difficulties
 to
 align
 different
 stamps
 used
 in
 subsequent
 process
 steps
 and
extreme
demands
on
surface
flatness
over
large
areas,
limit
its
applicability
on
a
larger
scale.





 12


From
 a
 scientific
 point
 of
 view,
 requirements
 on
 nanofabrication
methods
 are
often
quite
different,
compared
to
those
imposed
by
the
semiconductor
industry
or
other
present
or
future
large
scale
applications.
Flexibility
of
the
method,
i.e.
adjustability
 to
 different
 user
 needs,
 is
 often
 an
 important
 property
 since
 the
influence
of
different
nanostructure
properties
such
as
shape,
 size,
distribution
and
material
on
physical
properties
is
of
obvious
scientific
interest.
Examples
are
quantum
 effects
 in
 small
 particles,
 optical
 scattering
 by
 sub
 wavelength
structures
 or
 biological
 functionality
 of
 nanostructures.
 In
 these
 contexts
 the
pattern
replicating
techniques
are
not
very
suitable
since
often
only
small
series
of
 samples
 with
 identical
 properties
 need
 to
 be
 fabricated.
 The
 more
 flexible
techniques
like
EBL
or
STML
meet
the
requirement
that
structure
properties
can
easily
be
changed
but
are
on
the
other
hand
expensive
and
complex
methods.
For
patterning
of
surface
areas
of
several
cm2
they
are
also
slow.


Self‐assembly
methods,
 such
 as
 CL,
 are
 well
 fitted
 to
meet
 the
 demands
 for
 a
flexible,
affordable
and
large
surface
area
compatible
nanofabrication
technique.
Important
current
 limitations
of
 the
self‐assembly
based
methods
are
that
they
do
 not
 offer
 precise
 spatial
 positioning
 of
 structures,
 which
 hinders
synchronization
 of
 sequential
 process
 steps,
 and
 some
 lack
 of
 versatility
regarding
shapes
of
the
structures
that
can
be
fabricated.


Since
each
nanofabrication
technique
is
associated
with
different
advantages
and
limitations
it
is
often
convenient
to
combine
two
or
more
techniques
to
achieve
the
 desired
 nanostructure
 patterns.
 For
 example,
 PL
 can
 be
 used
 to
 define
structures
on
larger
length
scales,
such
as
electrodes,
while
CL
is
used
to
define
nanopatterns
in
the
regions
between
the
electrodes.
CL
can
also
conveniently
be
used
 to
 make
 guiding
 studies
 to
 find
 relevant
 sample
 parameters,
 like
nanostructure
 size,
 shape
 and
 inter‐particle
 spacing.
 Once
 the
 parameter
intervals,
where
the
nanostructures
exhibit
the
most
interesting
properties
for
a
certain
application,
 are
 identified,
more
costly
and
 time‐consuming
 techniques,
like
 EBL,
 can
 be
 used
 to
 fine‐tune
 the
 nanostructure
 geometries
 and
distributions
within
the
relevant
regimes.




 13


3 Hole‐mask
colloidal
lithography
(HCL)
The
 nanofabrication
 technique
 described
 in
 this
 thesis,
 hole‐mask
 colloidal
lithography
(HCL),
is
a
variation
of
colloidal
lithography
and
relies
on
techniques
that
are
already
frequently
used
in
other
nanofabrication
methods:


• Spin
coating
of
polymer
films
• Self‐assembly
of
colloidal
spheres
(nanospheres)
• Thin
film
deposition
• Reactive
Ion
Etching
(RIE)


In
this
chapter
the
focus
is
entirely
on
the
HCL
technique,
which
is
described
and
discussed
in
detail.



3.1 Colloidal
lithography
Colloidal
lithography
(CL)
is
currently
used
in
many
different
versions,
each
with
its
 own
 specific
 advantages
 and
 limitations[38,
 40].
 A
 frequently
 used
 version,
developed
by
van
Duyne
et.al.,
has
been
named
nanosphere
lithography
(NSL).
It
uses
 nanospheres
 in
 hexagonally
 close
 packed
 monolayers
 as
 etch
 or
evaporation
masks[43].
Since
the
nanospheres
are
arranged
on
the
surface
in
a
close
 packed
 crystal
 pattern,
 structures
 fabricated
 with
 this
 method
 are
 also
arranged
in
long
range
ordered
patterns,
i.e.
forming
a
lattice.
Spherical
particles,
even
when
packed
as
closely
as
possible
leave
gaps
in
between
adjacent
entities.
For
 the
 case
 of
 close
 packed
 spheres,
 these
 gaps
 have
 triangular
 shapes,
 with
each
 side
 constituted
 by
 a
 circle
 segment.
 In
 its
 simplest
 form,
 NSL
 uses
 the
nanosphere
monolayer
crystal
as
an
evaporation
mask,
 thus
 forming
triangular
nanostructures
 replicating
 the
 gaps
 in
 between
 the
 spheres.
 Several
 simple
variations
of
the
fabrication
process,
such
as
tilting
and
rotating
the
sample,
have
been
demonstrated
to
significantly
alter
the
shape,
separation
and
arrangement
of
the
nanostructures
that
can
be
fabricated
using
this
method[44].



3.2 Sparse
colloidal
lithography
In
 an
 alternative
 method,
 which
 will
 be
 referred
 to
 here
 as
 sparse
 colloidal
lithography
 (SCL),
 nanospheres
 are
 dispersed
 on
 surfaces
 from
 colloidal
solutions,
 not
 in
 a
 close
 packed
 pattern
 but
 in
 a
 sparse
 monolayer,
 i.e.
 with
separation
between
the
individual
spheres.
This
method,
previously
described
in
detail
 in
 several
 publications,
 has
 up
 to
 now
mainly
 been
 used
 to
 define
 etch
masks
for
ion
milling
processing[41,
45].
Below,
the
SCL
method
in
its
simplest
form,
producing
supported
nanodiscs
with
well‐defined
diameter,
thickness
and
average
separation,
will
be
reviewed.




 14


First
a
thin
film
of
the
material
of
choice
for
the
nanostructures
is
deposited
onto
a
flat
surface.
Already
here,
the
nanostructure
thickness
is
determined.
The
thin
film
is
then
covered
with
an
adhesive
layer
of
molecular
thickness,
consisting
of
a
polyelectrolyte
film,
onto
which
a
sparse
layer
of
nanospheres
is
adsorbed.
The
polyelectrolyte
film
is
used
to
provide
the
surface
with
a
charge
state
opposite
to
that
 of
 the
 colloidal
 particles,
which
 in
 turn
 facilitates
 their
 adsorption
 on
 the
substrate.
 Ion
 milling,
 using
 high‐energy
 ions,
 is
 then
 used
 to
 transfer
 the
nanosphere
 pattern
 into
 the
 thin
 film
 initially
 deposited
 on
 the
 surface.
 The
nanospheres
 act
 as
 a
 protective
 etch
mask,
 so
 that
 only
 the
material
 between
them
 is
 etched
 away.
 Residues
 of
 the
 spheres
 are
 finally
 removed
 using
 a
reactive
oxygen
treatment
(UV‐ozone,
or
oxygen
RIE).
By
changing
the
order
in
which
 the
 process
 steps
 are
 performed
 and
 by
 varying
 the
 process
 conditions
during
film
deposition
and/or
ion
milling,
e.g.
the
etch‐
or
evaporation
angle,
this
method
can
easily
be
used
to
fabricate
alternative
structures
such
as
nanorings,
crescent
or
extended
films
with
nanoholes[42,
46,
47].
The
development
of
hole‐mask
colloidal
lithography
(HCL)
described
below,
is
based
on
the
SCL
technique
and
 can
 be
 seen
 as
 an
 extension
 of
 SCL
 to
 increase
 its
 versatility
 and
applicability.



 15





3.3 Hole‐mask
colloidal
lithography
The
HCL
technique,
described
in
Paper
I,
uses
a
sacrificial
 layer
to
separate
the
nanosphere
mask
 from
 the
 surface
 to
 be
 patterned,
 which
 gives
 rise
 to
many
advantages
and
possibilities.
The
lithography
can
be
thought
of
as
composed
of
three
 major
 process
 steps,
 i)
 fabrication
 of
 a
 supported,
 patterned
 mask,
 ii)
transfer
of
the
mask
pattern
through
etching
and
iii)
transfer
of
the
mask
pattern
through
material
deposition.




3.3.1 Fabrication
of
a
supported,
patterned
mask





Figure
3.1
Schematic
description
of
mask
fabrication:
1)
PS
nanospheres
supported
on
a)
a
polymer
film
or
b)
directly
on
the
substrate
surface
are
used
as
an
evaporation
mask.
In
b)
the
sacrificial
layer
is
deposited
subsequent
to
the
polystyrene
spheres.
2)
Depending
on
deposition
angle
the
resulting
holes
in
the
deposited
mask
are
a
&
d)
round
and
replicate
the
sphere
diameter,
b)
elliptical
with
the
long
axis
larger
than
the
sphere
diameter
or,
c)
round
or
elliptical
with
diameter
smaller
than
the
nanospheres.
3)
The
nanospheres
are
removed
 using
 tape
 stripping
 or
 ultrasonic
 cleaning.
 4)
 The
 final
 result
 is
 a
 holemask
supported
on
a
sacrificial
layer.



 16



Figure
 3.2
 SEM
 images
 of
 a)
 Aumask
 supported
 on
 a
 polymer
 film,
 deposited
 from
 an
angle
 45°
 from
 the
 surface
 normal,
 before
 tape
 stripping
 away
 the
 190
 nm
 diameter
nanospheres.
The
holes
in
the
mask,
appearing
black
in
the
image,
are
partly
covered
by
the
 remaining
 (white)
 spheres.
 The
 shape
 of
 the
 holes
 is
 clearly
 elliptical.
 b)
 Aumask
resulting
 from
 sequential
 deposition
 from
 two
 opposite
 angles
 60°
 from
 the
 surface
normal,
after
tape
stripping.
The
holes
(black)
are
slightly
elliptical
and
with
roughly
half
of
 the
 diameter
 of
 the
 used
 110
 nm
 nanospheres.
 The
 elliptical
 grey
 areas
 represent
regions
where
the
mask
is
thinner
than
in
the
bright
grey
areas,
due
to
shadowing
of
the
evaporation
from
one
of
the
angles.
Occasionally
shadows
deriving
from
evaporation
from
opposite
angles
overlap,
which
shows
up
as
darker
grey
areas
(e.g.
near
the
centre
of
the
image).



Fig.3.1,
 Step
 1:
 a)
 The
 initial
 process
 step,
 is
 deposition
 of
 a
 sacrificial
 layer,
conveniently
 achieved
 by
 spincoating
 a
 thin
 polymer
 film
 onto
 the
 surface,
 a
process
 already
 well
 established
 and
 used
 for
 PL
 and
 EBL
 processing.
 The
polymer
 film
 is
 briefly
 treated
 in
 an
 oxygen
 plasma
 (5
 s.,
 50
W,
 250
 mTorr),
which
decreases
the
hydrophobicity
of
the
polymer
surface.
This
is
important
in
order
 to
 avoid
 spontaneous
 de‐wetting
 of
 the
 surface
 during
 subsequent
deposition
of
polyelectrolyte
and
nanospheres,
which
in
turn
is
important
to
get
a
high
quality,
homogeneous
surface
distribution
of
the
spheres.
A
water
solution
containing
 a
 positively
 charged
 polyelectrolyte
 is
 pipetted
 or
 poured
 onto
 the
polymer
 film.
 The
 next
 process
 step
 is
 deposition
 of
 negatively
 charged
nanospheres
 onto
 the
 adhesive,
 electrolyte‐covered
 polymer
 surface
 and
subsequent
 drying
under
 an
 intense
N2‐gas
 jet.
 To
 increase
 the
 stability
 of
 the
adsorbed
pattern
of
nanospheres,
the
samples
can
be
dipped
in
a
hot
fluid,
which
promotes
 adhesion
 between
 the
 colloids
 and
 the
 polyelectrolyte
 and
 polymer
film[48].


b)
 Alternatively,
 the
 nanospheres
 can
 be
 deposited
 directly
 onto
 the
 substrate
surface,
in
which
case
a
thicker
(triple)
polyelectrolyte
layer
should
be
used[48].
The
sacrificial
layer
is
then
evaporated
onto
the
surface,
prior
to
the
mask
layer
deposition.
A
 convenient
 choice
 of
 sacrificial
 layer
 for
many
 applications
 is
 Cr,
but
 other
 materials
 can
 of
 course
 also
 be
 useful.
 The
 main
 concern
 is
 that
 it
should
 be
 possible
 to
 selectively
 disolve
 the
 sacrificial
 layer,
 leaving
 the
substrate
and
the
fabricated
nanostructures
unchanged.

Fig.3.1,
 Step
 2:
 In
 the
 following
 step
 a
 thin
 film,
 resistant
 to
 oxygen
 plasma,
 is
deposited
 onto
 the
 surface.
 This
 layer
 is
 referred
 to
 as
 the
 hole‐mask
 and
depending
on
deposition
angle,
it
is
possible
to
control
the
shape
and
size
of
the
holes
in
the
mask
as
demonstrated
in
fig.3.2.
A
requirement
on
the
deposited
film
is
that
it
is
thinner
than
about
half
the
nanosphere
diameter
and
thick
enough
to
be
continuous.




 17



Fig.3.1,
 Step
 3‐4:
 Subsequently,
 the
 nanospheres
 can
 be
 removed
 by
 simply
attaching
 a
 piece
 of
 tape
 to
 the
 surface.
 As
 the
 tape
 is
 removed
 the
 spheres,
sticking
 harder
 to
 the
 tape
 than
 to
 the
 polymer
 surface,
 are
 removed
 as
 well.
Alternatively
 the
 nanospheres
 can
 be
 removed
 by
 cleaning
 the
 sample
 in
 an
ultrasonic
bath
and
 iso‐propanol.
At
 this
stage
the
polymer
supported
thin
 film
has
 holes
 in
 it
 where
 the
 spheres
 were
 covering
 the
 surface
 during
 the
deposition
process.
The
number
density
and
diameter
of
these
holes
correspond
to
the
shadows
of
the
nanospheres
on
the
polymer
film.
By
the
salt
concentration
and
 size
 of
 the
 nanospheres
 in
 the
 deposited
 colloidal
 solution,
 the
 number
density
and
diameter
of
the
holes
in
the
mask
can
easily
be
tuned[48].
Changing
the
angle
from
which
the
thin
film
is
deposited
can
also
alter
the
shape
and
size
of
 the
holes
 (fig.3.2).
Deposition
 from
any
angle
other
 than
 the
 surface
normal
results
 in
 stretched
 out
 shadows
 and
 thus
 elliptical
 holes
 in
 the
mask.
 On
 the
other
hand,
if
two
or
more
opposite
angles
are
applied,
material
is
deposited
in
under
 the
nanospheres,
 resulting
 in
 holes
 in
 the
mask,
which
 are
 smaller
 than
the
sphere
diameters.


3.3.2 Transfer
of
the
mask
pattern
through
reactive
ion
etching





Figure
 3.3
 Pattern
 transfer
 into
 the
 sacrificial
 polymer
 layer
 using
 oxygen
 plasma.
 The
duration
of
the
patterntransferring
plasmaetch
determines
the
degree
of
undercut.
5)
A
short
oxygen
plasma
treatment
gives
a)
little
or
no
undercut
and
b)
extended
etching
gives
a
controlled
undercut.
c)
For
the
example
with
a
Cr
sacrificial
layer
a
wetetch
in
Cretch
is
required
to
achieve
undercut
and
a
good
liftoff
 later
on.
6)
A
second
etch
process
can
be
applied
to
further
a)
transfer
the
pattern
into
the
substrate
or
b)
to
remove
the
holemask
prior
to
further
processing.



 18




Figure
3.4
a)
Size
of
the
holes
in
an
undercut
polymer
film
as
a
function
of
applied
oxygen
etch
 time
 starting
with
190
nm
holes
 in
 the
mask.
The
 SEM
 images
display
80
nm
 thick
polymer
 films
after
oxygen
RIE
 (50
W,
250
mTorr)
and
subsequent
removal
of
 the
holemask,
 corresponding
 to
 the
 longest
 b)
 (90
 s)
 and
 the
 shortest
 c)
 (40
 s)
 etch
 times
presented
in
the
graph.

Fig.3.3,
 Step
 5:
 The
 subsequent
 process
 steps
 aim
 at
 transferring
 the
 thin
 film
hole‐pattern
into
the
sacrificial
layer.
When
a
polymer
film
is
used,
this
is
easily
achieved
using
oxygen
RIE.
The
plasma
conditions
 can
be
 chosen
 so
 that
 all
 of
the
 polymer
 exposed
 under
 the
 mask
 holes
 is
 removed
 while
 the
 polymer
covered
by
the
film
is
unaffected.
Due
to
the
directionality
of
the
RIE
process
the
polymer
film
is
etched
predominantly
in
the
forward
direction.
Once
the
polymer
is
 completely
 etched
 through
vertically,
 the
 etching
will
 continue
 in
 the
 lateral
direction
 thus
 creating
 an
 undercut
 into
 the
 polymer
 film.
 The
 degree
 of
undercut
varies
linearly
with
applied
etch
time
and
can
be
controlled
to
within
a
few
 nanometers.
 For
 the
 Cr‐film
 sacrificial
 layer,
 on
 the
 other
 hand,
 the
 hole
pattern
 already
 extends
 all
 the
 way
 down
 to
 the
 substrate.
 However,
 a
 slight
undercut
 is
 always
 beneficial
 for
 subsequent
 material
 deposition
 and
 liftoff
processing.
A
suitable
undercut
can
be
achieved
by
wet‐etching
 in
a
Cr‐etchant
(e.g.
10
s.
in
Nickel‐Chrome
etch,
711.21
Sunchem
electrograde
products).





 19


Fig.3.3,
 Step
 6:
 If
 the
 hole‐mask
 is
 selectively
 removed
 prior
 to
 further
processing,
the
polymer
film
alone
will
constitute
the
lithographic
mask
and
the
undercut
 can
 be
 used
 to
 continuously
 increase
 the
 structure
 diameter.
 This
 is
demonstrated
 in
 fig.3.4
where
a
80
nm
thick
polymer
 film
covered
with
a
gold
hole‐mask
has
been
etched
in
an
oxygen
plasma
at
50
W,
250
mTorr
for
different
times.
After
pattern
transfer
into
the
sacrificial
layer,
the
hole‐mask
pattern
can
be
transferred
further
into
the
surface
by
choosing
the
proper
etching
conditions
for
the
surface
at
hand.
An
example,
which
has
been
demonstrated
in
Paper
I,
is
that
of
etching
into
a
TiO2
surface
using
CF4
RIE.
Care
has
to
be
taken
so
that
the
thin
 film
mask
material
 is
 resistant
 to
both
of
 the
etching
processes
applied
 to
penetrate
the
sacrificial
layer
and
to
extend
the
etching
into
the
substrate.



 20








3.3.3 Transfer
of
the
mask
pattern
through
deposition





Figure
3.5
Applying
different
deposition
parameters,
many
different
structure
shapes
and
distributions
can
be
obtained.
7)
a)
Deposition
of
a
thick
layer
of
nanostructure
material
results
 in
cone
shaped
 features.
b)
Using
 two
different
polar
angles
 to
deposit
materials
yields
 particle
 pairs.
 c)
 Deposition
 of
 materials
 through
 a
 mask
 where
 the
 pattern
 has
already
 been
 transferred
 into
 the
 substrate
 surface
 (by
 etching)
 results
 in
 nanodiscs
buried
 into
 the
etch
pits.
d)
Evaporation
 through
a
mask
where
 the
 top
 layer
holemask
has
 been
 selectively
 removed
 (without
 significantly
 affecting
 the
 polymer
 mask)
 gives
structures
 with
 a
 diameter
 replicating
 the
 undercut
 in
 the
 polymer
 film.
 e)
 Deposition
through
 the
 Crsupported
 holemask
 gives
 particles
with
 similar
 diameters
 as
 the
 holemask.
 8)
 Liftoff
 is
 achieved
 by
 immersing
 the
 sample
 in
 a
 solution,
 suitable
 for
 the
particular
sacrificial
layer.



 21



Figure
3.6
SEM
images
illustrating
some
of
the
different
structures
fabricated
with
HCL.
a)
and
b)
show
particle
pairs
with
different
separations;
 just
overlapping
and
separated
by
~10
 nm
 respectively.
 These
 structures
 were
 fabricated
 using
 identical
 deposition
conditions
and
holemasks
while
different
polymer
film
thicknesses
were
used
to
alter
the
particle
 separation.
 c)
Particle
pairs
made
up
of
 two
different
materials
 (Au
and
Ag).
 d)
Nanodiscs
in
holes
etched
into
the
surface.
e)
Elliptical,
layered
structures
of
Au
on
top
of
SiO2.
 f)
 Nanocone
 array.
 g)
 Nanocones
 and
 mask
 after
 partial
 liftoff.
 h)
 Inverted
 ring
structure,
Cr
on
Si.






 22


Fig.3.5,
 Step
 7:
 The
 process
 steps
 described
 so
 far
 have
 resulted
 in
 a
 thin
 film
hole‐mask,
 supported
 on
 a
 sacrificial
 layer
with
 a
 similar
 pattern.
 This
 double
layer
 mask
 can
 then
 be
 conveniently
 used
 as
 an
 evaporation
 mask
 to
 yield
 a
variety
 of
 nanofeatures
 as
 demonstrated
 in
 fig.3.5
 &
 fig.3.6.
 Collimated
deposition
 of
 materials
 from
 a
 source
 positioned
 along
 the
 surface
 normal
results
in
structures
on
the
surface
with
a
shape
replicating
the
holes
in
the
thin
film
 mask
 (fig.3.5e).
 If
 the
 hole‐pattern
 has
 already
 been
 transferred
 into
 the
surface,
 the
 structures
 will
 end
 up
 at
 the
 bottom
 of
 these
 holes
 (fig.3.5c
 &
fig.3.6d).
 In
 this
 way
 nanostructures
 can
 be
 incorporated
 into
 a
 surface
 film
rather
than
just
placed
on
top
of
it.


Deposition
 of
 materials
 from
 a
 source
 positioned
 anywhere
 off
 the
 surface
normal
results
in
structures
laterally
displaced
with
respect
to
the
centre
of
the
mask
 hole.
 Choosing
 two
 opposite
 polar
 angles
 to
 deposit
materials
 from
 thus
results
 in
 a
pair
 of
 structures
 (fig.3.5b).
Both
 the
 thickness
of
 the
double
 layer
mask
and
 the
deposition
angles
can
be
used
 to
control
 the
separation
between
the
two
structures
in
the
pair.
 It
 is
however
preferable
to
alter
the
thickness
of
the
mask
 and
 keep
deposition
 angles
 constant
 to
 avoid
 shape
 aberrations
 that
may
 be
 introduced
 by
 deposition
 from
 steep
 angles
 (>20°
 from
 the
 surface
normal).
 Using
 the
 same
 strategy
 it
 is
 also
 possible
 to
 fabricate
 three
 or
more
structures
from
each
hole
in
the
mask.


If
 thick
 layers
 are
 deposited
 through
 the
 mask
 the
 structures
 growing
 on
 the
surface
will
gradually
attain
a
noticeably
smaller
diameter.
This
is
related
to
the
deposition
of
materials
on
top
of
and
on
the
rims
of
the
mask‐holes,
which
tend
to
gradually
decrease
the
diameter
of
the
holes.
When
depositing
particle
pairs,
this
has
the
consequence
that
if
the
different
particles
corresponding
to
different
deposition
 angles
 are
 fabricated
 sequentially,
 the
 first
 structures
 will
 attain
 a
larger
 diameter
 than
 the
 second
 ones
 (fig.3.6c).
 This
 can
 be
 avoided
 by
frequently
 altering
 between
 the
 two
 chosen
 deposition
 angles
 (fig.3.6a
 &
 b).
Another
 consequence
 of
 the
 shrinking
mask‐holes
 is
 that
 extended
 deposition
through
 circular
 holes
 in
 the
 mask
 result
 in
 cone‐shaped
 structures
 on
 the
surface
 (fig.3.6f
 &
 g).
 The
 rate
 at
 which
 the
 holes
 close
 is
 dependent
 on
 the
deposited
 material,
 which
 means
 that
 thick
 structures
 composed
 of
 different
materials,
will
have
a
different
side‐wall
angle
at
different
parts
of
the
cone.
This
can
be
used
to
place
discs
of
different
sizes
on
top
of
each
other,
separated
by
a
spacer
layer[27].


If
 the
mask
material
 is
 properly
 chosen
 it
 can
 be
 selectively
 removed
 prior
 to
material
 deposition
 (using
 for
 example
wet
 etching),
 leaving
 only
 the
 polymer
mask
on
the
surface.
This
has
the
advantage
that
the
under‐etch
of
the
polymer
then
can
be
used
 to
 fine‐tune
 the
diameter
of
 the
structures
 that
are
produced
(fig.3.5d).
Another
example
where
the
hole‐mask
rather
than
the
sacrificial
layer
is
 removed
 is
 demonstrated
 with
 the
 Cr
 sacrificial
 layers.
 Deposition
 of
 Cr
through
 a
 Au
 hole‐mask,
 after
 a
 short
 under‐etching
 into
 the
 Cr‐layer
 and
removal
 of
 the
 hole‐mask
 using
 a
 suitable
wet
 etchant
 (Au‐etch),
 results
 in
 an
inverted
Cr
 ring
 structure
 (fig.3.6.h)
 (or
 equivalently,
 Cr
discs
 in
holes
 in
 a
Cr‐film).
This
fabrication
strategy
can
of
course
be
altered
to
include
two,
between
themselves
different
materials,
e.g.
Pt
discs
in
a
Cr
hole‐film.





 23



Fig3.5,
Step
8:
In
the
last
process
step
the
sacrificial
layer
can
be
removed
using
an
appropriate
solvent.
Usually
acetone
works
well
for
polymers
and
Cr‐etch
for
Cr‐
sacrificial
layers.
It
should
be
pointed
out
that
the
thickness
of
the
sacrificial
layer
 film
 should
 be
 chosen
 to
 be
 comparable
 to
 or
 thicker
 than
 the
nanostructures
 intended
 for
 fabrication.
 With
 a
 too
 thin
 sacrificial
 layer,
nanostructures
 deposited
 through
 the
mask
will
 stick
 up
 through
 and
 into
 the
hole‐mask,
 which
 may
 result
 in
 partial
 removal
 of
 the
 structures
 due
 to
establishment
of
physical
contact
with
the
hole‐mask.



3.3.4 Limitations
and
extensions
The
 above
 described
 method,
 with
 a
 polymer
 sacrificial
 layer
 works
 well
 for
fabrication
of
discs,
ellipses,
particle
pairs
or
cones
of
any
material
 that
can
be
deposited
from
a
reasonably
collimated
source,
on
most
surfaces
flat
enough
to
spin
 coat
 and
 that
 are
 not
 very
 sensitive
 to
 oxygen
 RIE.
 A
 great
 advantage
 of
always
 using
 the
 same
 sacrificial
 layer
 material
 is
 that
 the
 deposition
 of
nanospheres
onto
the
surface
is
standardized.
Therefore,
the
deposition
process
will
not
suffer
from
variations
due
to
intrinsic
properties
of
the
surface
and
thus
will
 be
 substrate
 independent.
 Using
 patterned
 polymer
 masks
 also
 has
 the
advantage
that
it
is
a
well‐established
technique,
already
used
for
a
long
time
in
EBL
and
PL,
which
means
 that
deposition
and
 liftoff
procedures
 can
 simply
be
adopted
 from
 existing
 processes.
 In
 addition,
 polymer
 liftoff
 can
 be
 done
 in
acetone
or
 similar
mild
 solvents,
which
 is
 compatible
with
most
nanostructure
materials
 and
 surfaces.
 Another
 obvious
 advantage
 of
 the
method
 is
 that,
 like
other
 self‐assembly
based
 techniques,
 it
 is
 a
 parallel
 process
 and
 thus
 suitable
for
patterning
of
large
surface
areas.


Although
polymer
films
are
very
versatile
as
sacrificial
layers,
certain
situations
may
 benefit
 from
 the
 use
 of
 other
 materials.
 For
 example,
 fabricating
nanostructures
 on
 a
 surface
 that
 is
 oxidized
 and
 etched
 requires
 the
 use
 of
 a
sacrificial
 layer
 that
 can
 be
 etched
without
 the
 aid
 of
 oxygen
 plasma.
 Another
situation
 where
 Cr‐films
 are
 preferred
 over
 polymers
 is
 when
 the
 masks
 are
exposed
 to
 high
 temperature
 and/or
 intense
 irradiation
 prior
 to
 liftoff,
 since
such
conditions
can
damage
the
polymer
and
prevent
 liftoff.
As
usual,
 the
hole‐mask
has
to
be
chosen
to
resist
the
sacrificial
layer
etchant.
For
Cr‐films,
Au
hole‐masks
fulfils
this
criterion.



A
 limitation
 of
 the
method,
 at
 present,
 is
 the
 lack
 of
 long‐range
 order
 and
 the
limits
 in
 particle
 separation
 that
 can
 be
 achieved.
 Long‐range
 ordering
 can
 be
achieved
 for
 example
 with
 templated
 deposition
 or
 printing
methods,
 but
 the
lack
of
ordering
can
also
be
an
advantage.
Since
the
structures
are
not
ordered
in
crystal
like
patterns
collective
effects
like
optical
interference
tend
to
cancel
out.
The
 typical
 nanosphere
 coverage
 that
 can
 be
 achieved
 with
 electrostatic
 self‐assembly
 has
 been
 shown
 to
 range
 between
 12
 and
 52
 %
 (projected
 surface
area)[48].
 Although
 the
 upper
 limit
 is
 not
 easily
 overcome
 without
 causing
agglomeration
 of
 the
 particles,
 the
 lower
 limit
 can
 be
 pushed
 by
 some
 simple
tricks.
 Starting
 with
 larger
 spheres
 and
 shrinking
 the
 holes
 in
 the
 mask
 by
deposition
 from
 off‐normal
 angles
 is
 one
 way
 to
 decrease
 surface
 coverage.



 24


Additional
to
this,
the
nanospheres
can
be
treated
in
oxygen
plasma,
for
a
short
time,
prior
 to
hole‐mask
deposition,
 to
 further
shrink
 the
structure
sizes.
Mask
holes
with
 diameters
 smaller
 than
 the
 nanospheres
 is
 also
 possible
 to
 achieve
from
sputter
deposition
of
the
hole‐mask[49].


The
size
dispersion
of
the
fabricated
nanostructures
is
largely
determined
by
the
size
dispersion
of
the
masking
nanosperes,
which
 is
generally
 in
the
5
%
range
for
spheres
with
average
diameters
of
50
nm
and
larger.
For
colloidal
solutions
containing
 smaller
 particles
 (commercially
 available),
 size
 dispersions
 are
considerably
worse
 (10
%
 or
more)
 and
 thus
 set
 a
 lower
 limit
 for
 the
 size
 of
structures
 that
 can
 be
 fabricated
 with
 reasonable
 conformity
 of
 the
 achieved
diameters.
 However,
 this
 applies
 to
 commercially
 available
 colloidal
 solutions,
and
 for
 applications
 where
 narrower
 size
 distribution
 are
 crucial,
 this
 can
 be
achieved
e.g.
by
chromatographic
methods.

Furthermore,
 it
 is
 worth
 mentioning
 that
 preparation
 of
 nanostructures
 via
evaporation
of
materials
through
a
supported
mask
depends
on
the
collimation
of
 the
 atomic
 beam.
 In
 deposition
 techniques
 as
 sputtering,
 chemical
 vapor
deposition
 or
 laser
 ablation,
 which
 operate
 at
 higher
 pressure,
 the
 deposited
atoms
 are
 frequently
 scattered
 by
 the
 process
 gases
 resulting
 in
 poor
collimation.
 This
 results
 in
 nanostructures
 with
 ill‐defined,
 blurry
 edges.
 In
addition,
due
to
 the
anisotropic
deposition
conditions,
materials
are
deposit
on
vertical
 walls
 as
 well
 as
 on
 the
 horizontal
 surfaces.
 Under
 such
 conditions,
 a
continuous
 film,
 connecting
 the
 mask
 and
 the
 nanostructures
 can
 be
 formed,
which
in
turn
can
result
in
poor
liftoff
performance.

 


3.4 Applications
of
HCL
to
fabricate
carbon
nanostructures
This
 section
 describes
 the
 specific
 application
 of
 HCL
 to
 the
 fabrication
 of
nanostructured
carbon
materials
as
demonstrated
in
Papers
II‐V.
Two
different
strategies
where
 adopted,
 one
 for
 etching
 out
 carbon
nanostuctures
 from
bulk
carbon
 surfaces
 and
 another
 for
 fabrication
 of
 amorphous
 carbon
nanostructures
 on
 various
 surfaces.
 The
 former
 technique
 utilizes
 HCL
 to
prepare
etch
masks
on
carbon
surfaces
and
subsequent
oxygen
RIE
to
define
the
nanostructures,
whereas
the
latter
employs
a
sacrificial
Cr‐layer
and
subsequent
e‐beam
 evaporation
 of
 carbon.
 Details
 and
 considerations
 related
 to
 the
 two
fabrication
processes
are
discussed
below.



 25






3.4.1 Carbon
nanostructures
from
bulk
materials
One
of
the
benefits
of
using
bulk
materials
and
etch
out
the
nanostructures
from
these
 is
 that
 the
 structures
 inherit
 the
 physical
 properties
 of
 the
 original
material.
Using
this
approach
it
is
therefore
possible
to
fabricate,
e.g.
crystalline
graphite
nanostructures.
To
achieve
 this,
 the
 specific
 strategy
employed
 in
 this
work
 (Paper
 II‐IV)
was
 to
 first
 fabricate
 gold
 nanodiscs
 on
 the
 carbon
 surface
using
HCL,
and
then
to
use
the
gold
discs
as
etch
masks
 in
an
extended
oxygen
RIE
 process
 to
 transfer
 the
 pattern
 onto
 the
 underlying
 surface.
 The
manufacturing
of
gold
discs
on
the
carbon
surface
can
be
performed
according
to
the
processes
described
in
the
previous
sections,
with
some
minor
differences
in
some
of
the
process
steps.


After
 spin
 coating,
 deposition
 of
 nanospheres
 and
 deposition
 of
 the
 hole‐mask
film
as
usual
(fig.3.1,
step1
&
2).
 the
nanospheres
needs
to
be
removed.
This
 is
preferably
done
using
a
mild
ultrasonic
cleaning
in
IPA.
Particularly
for
graphite
this
method
 is
preferred
over
 tape
stripping
since
 the
bonds
between
adjacent
planes
 of
 a
 graphite
 crystal
 are
 very
 weak.
 Tape
 stripping,
 exerting
 forces
perpendicular
 to
 the
 surface,
 risks
 to
 break
 the
 bonds
 between
 the
 graphite
sheets,
 rather
 than
 between
 the
 nanospheres
 and
 the
 polymer
 surface,
 thus
lifting
off
the
uppermost
surface
layers
together
with
the
mask
instead
of
just
the
spheres.


Another
 step
 that
 requires
 extra
 attention
 is
 pattern
 transfer
 into
 the
polymer
film
 (fig.3.3,
 step
 5)
 since
 the
 underlying
 carbon
 surface
 is
 also
 sensitive
 to
oxygen
plasma,
although
much
less
so
than
the
polymer.
Prolonged
oxygen
RIE
after
 complete
 penetration
 of
 the
 polymer
 film
 might
 lead
 to
 damage
 of
 the
carbon
 surface
 although
 such
 effects
 can
 be
 minimized
 by
 choosing
 a
 proper
plasma
etch
duration.
After
gold
deposition
and
liftoff,
 leaving
gold
discs
on
the
surface,
the
pattern
can
be
transferred
into
the
carbon
surface
via
oxygen
RIE.





 26


The
 response
 of
 carbon
materials
 to
 oxygen
 plasmas
 is
 in
 itself
 an
 interesting
and
complex
topic
that
has
been
thoroughly
investigated
in
the
past
and
will
be
discussed
further
in
Chapter
4[50‐54].
As
demonstrated
in
Paper
II,
the
response
to
oxygen
RIE
varies
for
different
carbon
materials.
To
exemplify,
the
etch
rates
in
a
50
W,
250
mTorr
plasma
(generated
in
the
system
described
in
Chapter
6)
are
 given
 for
 the
 three
 carbon
based
materials
PMMA,
GC
and
HOPG.
The
etch
rate
in
the
lateral
direction,
causing
etching
under
a
mask
supported
on
a
80
nm
PMMA
film
is
1.25
nm/s
(derived
from
the
graph
in
fig.3.4).
In
GC
and
HOPG
the
corresponding
values
are
0.15
and
less
than
0.015
nm/s
respectively.
It
has
to
be
noted
that
the
comparison
between
etch
rates
in
the
polymer
film
and
the
pure
carbon
materials
 is
 somewhat
 unfair
 since
 the
 former
 value
 is
 derived
 from
 a
situation
where
the
etching
has
already
reached
the
inert
substrate
surface
and
thus
 there
 is
 no
 forward
 etching.
 Under
 these
 conditions
 more
 oxygen
 is
available
for
etching
in
the
lateral
direction
and
the
value
of
the
etch
rate
in
the
lateral
 direction
 is
 overestimated.
 The
 relations
 between
 the
 etch
 rates
 in
 the
forward
 direction
 for
 the
 considered
materials
 are
 similar
 to
 the
 lateral
 rates,
although
 significantly
 higher
 in
 absolute
 numbers
 due
 to
 the
 more
 efficient
supply
of
oxygen.
For
the
PMMA
film
the
forward
etch
rate
is
more
than
2
nm/s
whereas
 corresponding
 values
 for
 GC
 and
 HOPG
 are
 0.65
 and
 0.19
 nm/s
respectively.



3.4.2 Evaporated
carbon
nanostructures
In
 many
 situations
 it
 is
 desirable
 to
 be
 able
 to
 choose
 different
 materials
 to
support
 the
 nanostructures.
 Measurements
 of
 optical
 properties,
 for
 example,
are
 simplified
by
 the
use
 of
 transparent
 substrates.
 For
 such
purposes,
 carbon
nanostructures
can
be
fabricated
using
the
standard
HCL
technique,
described
in
the
 previous
 section,
 on
 any
 substrate.
 However,
 since
 deposition
 of
 carbon
through
 evaporation
 requires
 very
 high
 temperatures
 of
 the
 carbon
 target
material,
 some
 problems
 may
 be
 experienced
 using
 polymer
 films
 as
 liftoff
layers.
High
temperatures,
accompanied
with
emission
of
UV
radiation
is
known
to
change
the
properties
of
polymers,
 inducing
cross‐linking,
which
 in
 turn
can
influence
 the
 possibility
 to
 dissolve
 the
 sacrificial
 layer.
 Even
 for
 cases
 where
liftoff
 is
successful,
unwanted
residues
can
be
observed
on
the
surface
(fig.3.7).
An
 alternative
 explanation
 for
 the
 observed
 residues
 is
 that
 carbon
 from
 the
evaporation
 process
 deposits
 on
 the
 polymer
 film
 side‐walls
 thus
 creating
 the
observed
residues.





 27



Figure
3.7
Amorphous
carbon
nanostructures
on
a
fused
silica
substrate.
The
sample
was
prepared
 by
 evaporation
 of
 carbon
 through
 a
 polymer
 sacrificial
 layer.
 Slow
 liftoff
 is
observed
as
well
as
residue
material
around
several
of
the
nanostructures.

Even
 though
 the
 effect
 with
 residue
 halos
 around
 the
 nanostructures
 can
potentially
 be
 exploited,
 it
 is
 necessary
 to
 be
 able
 to
 avoid
 it.
 Using
 the
 Cr‐sacrificial
layer
instead
of
polymers
circumvents
this
problem.
Cr
has
the
benefit
of
not
being
as
sensitive
to
high
temperatures
and
UV‐irradiation
as
polymers.
As
demonstrated
 in
 Paper
 V,
 this
 approach
 can
 be
 used
 to
 achieve
 large
 surface
areas
covered
with
amorphous
carbon
nanostructures
on
fused
silica.



3.5 Other
applications
of
HCL
As
demonstrated
in
this
thesis,
HCL
is
useful
for
fabrication
of
nanostructures
in
several
different
types
of
applications.
In
addition
to
these,
many
other
examples
can
be
found
in
a
large
variety
of
research
fields.
Optical
properties
of
cylindrical
nanostructures
in
Pt,
Pd,
Ag,
Au,
Al
have
been
prepared
for
studies
of
their
optical
properties
both
with
near‐
and
far‐
field
techniques[24,
26,
28,
55,
56].
Tri‐layer
sandwich
structures
in
Au‐SiO2‐Au
have
been
prepared
for
investigations
of
their
optical
and
magnetic
properties[27].

Furthermore,
Au,
Pd
and
Pt
nanostructures
have
been
fabricated
for
research
targeting
applications
in
optical
sensing,
hydrogen
storage
and
electrochemistry[25,
29,
30,
57].
These
applications
further
demonstrate
the
usefulness
of
the
developed
HCL
technique
and
many�